open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am trying to model a river bend and produce lateral erosion on the exterior bank and deposition on the interior bank. For that reason I am applying secondary currents in the T2D and the GAIA files.
However, the calculation always results in a very irregular riverbed, which does not look right (see screenshot).
I already refined the mesh and decreased the timestep so that the CFL criteria is smaller < 0.1 (Mesh triangle size = 0.25 m, vel_max = 2 m/s, TS = 0.01).
Changing the alpha coefficient of the secondary currents has only little effect. In order to see the effect of secondary currents I switched off the lateral bed slope option and slope failure for now.
Besides I compared my .cas files to the Yen bend example and I cannot find any obvious differences.
What is striking about the results, is that the irregular shape of the river bed somewhat follow lines of the semi irregular shaped mesh (see screenshot). Could this maybe be a reason? The mesh of the Yen example is completely irregular...

I attached the .cas files, the mesh file and a screenshot of the irregular river bed I am producing.

Has someone encountered similar issues or has an idead what might be the issue in my calculation?

Thanks for any advice in advance!

Best regards,
Jakob

I updated the .cas files to improve their readability.

Dear Jakob,
I took a quick look to your cas files.
Everything looks OK, but I’ve some few suggestions to propose. Your configuration is far too too ambitious, particularly to detect some possible source of errors/improvements :
- (1) Run only the hydrodynamics, take a look to the flow field to see if it’s OK. Print also the shear stress, and look at it. Don’t activate SC.
- (2) Run the same simulation, now activating SC, only hydrodynamics. Take a look particularly to the velocity field and shear stress distribution.
- (3) Call to GAIA for the case (1), use only one sediment class, don’t activate SC in GAIA nor sediment slide. Take a look to the result.
- (4) Run (3), and now activate SC in GAIA.
- (5) Run (4), and now activate sediment slide. BTW, 65 degrees seems a bit high non ?
- (6) Run (5), now activate SC in the hydrodynamics.
- (7) Run (6), by including progressively the different sediment classes.

For the simulation runs, I'd suggest to activate the slope effect. This can help to the stability of the system by adding (physical) diffusion.

Best wishes,
Pablo

Hello,

thank you for all the hints, they definetly helped to narrow down the error and I finally found it:
I had SUPG 0,0,2,2. I switched it to 2,2,2,2 and it looks way better now.

As a side note:
Secondary currents option alone still gives me a very irregular river bed. But with the consideration of slope effects, the river bed becomes more smooth. Visually, the best result is obtained with Formula for devtiation = Talmon and Slope effect = Koch & Flokstra, similar to the YEN example.

Thank you again and best regards,
Jakob